The Weatherman

[imoldernu]

Some thoughts, and wondering how others feel about weather forecasts.

Maybe my imagination, but it seems that the accuracy of forecasts this year are worse than ever... at least here in the midwest. Not just for the long term (as in one week) but even for tomorrow.

We look to the forecasts to plan for holidays, especially in the summer, when we have cookouts at our camp, which means that our friends have to drive an hour or so to get there. This year, the three day forecasts for Memorial Day, Fourth of July, and Labor day have been so wrong that weekends have either been ruined by unpredicted bad weather, or bad weather forecasts have forced families to cancel outdoor planned events, when the weather turned out to be excellent...

That's not the only problem... My current complaint is one that can affect hundreds of thousands of RV'ers, who have vehicles in the cold belt. It has to do with forecasts of freeze. At some point, RV's of all kinds require winterizing to protect water and sewer lines. In our case, it means blowing out water lines water lines and adding RV antifreeze. To have this done in our campgrounds costs from $50 to $100. Not doing this could result in a freeze that could cause as much as $1500 or more to repair.

Common knowledge has it that for RV pipes to freeze, requires a temperature below 28 degrees that last for four to eight hours (varies by source).

So here's the situation today is Thursday. Last night, the weather forecast for the Friday night was for a low of 34 degrees, and a daytime temp of 54 on Saturday. This morning the new forecast is for 26 degrees and 42 .

For most people that wouldn't make much of a difference, but for RV owners, many of whom, like me, would like to extend our season by a few weeks, it means dropping everything to drive to camp, and winterize. For those who hire to have it done, they're out of luck... too late.

Not to put too fine point on this, but looking at the longer term forecast... we are not scheduled to have a single day between now and November 20th, when the temperature is forecast to go below freezing for the first time. That's 3 weeks from today.

And yet, I'll still tune in to watch the giggly weather girl to read from the prompter... even though I know she has no idea...

We pay $4.900,000,000 (that's billions) for the National Weather Service that can't tell what the weather will be like tomorrow...

Curmudgeonly yours,
imoldernu


so back to the old standby:

 

[bbbamI]

I think they're smokin' doobies...

 

[eytonxav]

And I thought this post was about Mark Rudd

 

[GrayHare]

Would you rely on a "giggly health girl" for important medical information? For scientific info get a scientist's estimate, many of which can be found online.

 

[MichaelB]

.
.
.
Weather Prediction From Groundhog Day - YouTube

 

[Car-Guy]

Some thoughts, and wondering how others feel about weather forecasts.

Maybe my imagination, but it seems that the accuracy of forecasts this year are worse than ever... at least here in the midwest. Not just for the long term (as in one week) but even for tomorrow.

:

I thought it was just my imagination too. They still seem to get night and day right but I don't need them for that. These days it seems they are right about half the time (maybe a little more) but that's not very good.

 

[RockyMtn]

pretty much the only job you can be wrong every day and not get fired!

 

[Car-Guy]

pretty much the only job you can be wrong every day and not get fired!

I had to go back and re-read the OP. I thought for a minute you were talking about the IRS.

 

[ls99]

So what do you expect from all that high tech leading edge sit in the room analists?

They are at least as good as the farmer's almanack, where thay careully observe caterpillars, tree bark, color of deer hide and tea leaves.

The part that annoys me is that that on commercial TV the weather blather bloviator is probably the highest paid character. Yet most of them use the NOAA source data, and modify it slightly so it seems more accurate for the locality. While some run their own radars for current info their long range guesstimates are sourced from the gummint.

True there are a few independently run weather predictors, they seem no more accurate.

But their historic data is spectacularly accurate.

Recently NOAA complained of not getting current satellite data. Can't figure out if that is relevant or not.

Do recall w*rking in the Aleutians, listening to NOAA short wave reports from the various remote locations. The old and not bold pilots would make as good of a prediciton from listening to the raw data radioed in, as the official forecast on the weather frequencies. And for about as far in advance as "professionals".

Edit add: Maybe Nords will pipe up. IIRC his spouse was in the Navy weather bureau.

 

[ERD50]

Is there any actual historical reporting on accuracy? It does seem way off, I've tried scheduling yard work and fertilizer and weed control spreading around the reports, and I should have just gone with my gut.

-ERD50

 

[Walt34]

Maybe it's easier to forecast around here, but they (NOAA) seem remarkably accurate to me.

Of course they did blow it big time in 1966 when we got three feet of light flurries. But they didn't have satellites and probably not even radar.

That's the thing though. Everyone remembers the screw ups and not the times they got it right.

 

[Ronstar]

I fell imoldernu's pain, and I agree that weather forecasting in the midwest is a guess. It seems to me that forecasting in other parts of the US is more accurate than the Chicago area. I haven't taken my boat out of the lake yet - doing it saturday morning. I shouldn't have any freezing problems - the water will be warmer than the air.

 

[Rustic23]

What is accuracy? Is it the cold front came in on the day it was predicted, or the morning vs afternoon, or at 9:00 when they said it would. Did it rain in my area or somewhere close? Did the temperature reach 45 deg or was it closer to 50? Were the winds 35 mph or stronger, if weaker did we even notice? If my area had a tornado and it was not forecast, but it was forecast fifty miles away, is that inaccurate?

My point is, I have been depending on weather forecast since 1966 when I started flying. I remember the wrong ones but seldom give the right ones another thought.

I may be wrong on this part, so if there is a weatherman out there please correct me, but a conversation with an Air Force weatherman he told me that the weather service was practicing Newtonian Weather. Seems lime Isaac Newton said if you have enough data you can predict the outcome. I understand the weather bureau gathers infomation 23.75 hours a day and feeds it into a super computer. It then grinds for 15 min and makes the forecast. I think they may be rethinking Newton and going to chaos theory.

I can not think of any job in the world that could be harder. If you look at all the lows and high fronts in the world and they are all pushing and pulling at each other and you have to decide if the water temp in the Indian ocean will effect a front over Africa, that effects the temperature in Germany, which effects a front in Russia, which will effect the rain in Las Vegas at 9:00 Saturday.

It makes timing the stock market seem childs play.

 

[braumeister]

Where I live, I've been absolutely delighted with the continually increasing accuracy and precision of the weather forecasts I get online.

And I'm being perfectly serious about that.
In all fairness, forecasts here used to be abysmal, but they are so much better today that I'm truly impressed.

 

[easysurfer]

I kinda miss the good old days when the temperature got hot in summer and cold in winter.

At least now the weather folks earn their pay

 

[Bestwifeever]

Imold, you sound really frustrated. Try this:

Lookup your first and last freeze/frost dates by zip code

Plug in the zip code and you'll get information about freeze dates. I would disregard a forecast of 11/20 for the first frost anywhere of any year in Illinois.

 

[imoldernu]

Imold, you sound really frustrated. Try this:

Lookup your first and last freeze/frost dates by zip code

Plug in the zip code and you'll get information about freeze dates. I would disregard a forecast of 11/20 for the first frost anywhere of any year in Illinois.

Thanks for the link
If there's any frustration, it's because the TV weather people don't seem to have a clue that their forecasts actually affect peoples' lives.
My kids had planned to visit tomorrow and looking at the forecast on Wednesday, it looked for sure as if Friday would be relatively warm (at least above freezing). Now, 24 hours later, we have to cancel, as Ill have to go to my park model trailer and winterize to avoid a potential big repair bill. There are 6000+ trailers in our campground, and based on talking to some of the contractors, there are still about 2000 to be winterized.
In all likelihood, the cold won't last long enough to do a hard damaging freeze, but for those people who live 150 miles away, and haven't winterized yet, there will be some anxious moments. We watched a neighbor who had an Airstream, where the lines froze early, broke, and by the time they got to the trailer the water had risen to the window level. (Airstreams are pretty tight). $4000 damage.

The freeze /frost date chart is very good, and looks to be a key to long range planning, but for instance, the first 28 degree freeze here, is listed for November 6. When it comes to $$$ risk, averages don't count.

An imprecise science... good enough to predict % chance of precipitation, but nowhere a (+- ) estimate on the temperature. Small wonder that satellite launches have problems.

Umm... Have you ever heard the weather girl say "I'm sorry... we made a mistake"?
Oh well, now I feel better... though I'm sad to think I'll miss the next two weeks at camp.

 

[nuke_diver]

If you are listening to the TV weather guys that might be part of the problem . I believe forecasting is much more accurate than it used to be and is accurate further out as well. Still Mother Nature will do what she wants sometimes, so that front that is suppose come in and bring you rain falls apart 100miles out, or shifts 100miles north and misses you.

The best forecasts you can get IMO is to read the actual NWS detailed forecasts. There a lot of terms to learn , things like 500mb temps etc but you will get the best details this way. And what you can often find (especially for around here) is that the various models that exist can give different results, one hotter one colder one wetter one drier. If you listen to the news you can often figure out which model the weatherperson favours.

When all the models say the same thing and do it consistently for a few days the forecast confidence will be high, when they don't is won't

 

[Bestwifeever]

Yes, try weather.gov for more weather information than you can ever use . The TV news weather has to be embellished and dramatized.

But forecasts are always refined as we get closer to the time being forecast. Better to find out today that the temp is probably going to be worrisomely cold on Saturday than to wake up to it on Saturday when it hits.

 

[rbmrtn]

And I thought this post was about Mark Rudd

My first thought was that group also

Not too bad in my area. I think one issue is they give a forecast for their viewing area which can cover a good bit of territory. The weather in your specific location can be a lot different. Local Weather from AccuWeather.com - Superior Accuracy™ is pretty good.

 

[Fred123]

The National Weather Service is suffering from a severe shortage of forecasters (> 500 as of last May) due to budget cuts from sequestration over the last couple of years. They are also way behind both Canada and Europe in the computer hardware that's available for running the weather models. As someone who worked at NOAA for more than 15 years, I can also say that NOAA management is extremely dysfunctional -- much more interested in investing in layers of bureaucracy and PR projects (I know someone who is making > $130k/year + Fed benefits for producing YouTube PR videos) then investing in needed scientists/forecasters and infrastructure.

 

[GrayHare]

Is there any actual historical reporting on accuracy? It does seem way off, I've tried scheduling yard work and fertilizer and weed control spreading around the reports, and I should have just gone with my gut.

-ERD50

The 5-days-out accuracy has now reached where 3-days-out had been about 20 years ago. Accuracy records are kept, primarily to compare the results of the various numeric models that exist (the USA's, Europe's, Japan's, etc.) and are run on ultra-high-speed computers. In recent years the European model generally has had a slight accuracy edge over the USA's primary one. However, any human forecaster who simply goes by what a computer suggests rarely performs as well as one who knows the local tendencies.

 

[ERD50]

Imold, you sound really frustrated. Try this:

Lookup your first and last freeze/frost dates by zip code

Plug in the zip code and you'll get information about freeze dates. I would disregard a forecast of 11/20 for the first frost anywhere of any year in Illinois.

That site gives some good information. I get frustrated when I hear the weather reporters on TV/Radio say things in Spring like 'today is the average date of the last frost, so it should be safe to put your sensitive plants out now'.

No, there's roughly a 50/50 chance (the distribution is probably not normal) we will still have another frost. Why not tell us that, and also when there is a 90% chance of no more frosts?

-ERD50

 

[H2ODude]

Used to manage a drought threatened water supply, so paid a LOT of attention to forecasts, primarily precipitation related. I would print 10 day forecasts and tape them to the wall outside my office to demonstrate that anything more than 3-4 days in advance was pretty much a crapshoot and not to be trusted. It may have improved since then, but I doubt by much. The only thing longer range we gave any credence to was the El Nino/La Nina oscillation, and it wasn't something significant enough to do much more than instill fear into a water supply person's heart. Glad those days are over for me (part of my signature line!)

 

[Lsbcal]

An app I like is Yahoo weather. It shows data for hourly over the next 48 hours or so. It's very accurate the closer we get to the event time.

Like this morning when I want to go for a run. Last night it said there would be possible rain in the AM although it did give a probability of 30% which is a nice feature. Now at 8am it says in an hour the temperature will be OK to run. But my outside temperature gage and eyes are the ultimate decider.